To setup on Mac or Linux:

bash mac-setup.sh
source venv/bin/activate

To setup on Windows:

windows-setup.bat
venv\Scripts\activate.bat

The Streamlit tests contain most tests of basic functionality.

Run

streamlit run streamlit_test.py --server.port 8555

And then run

npm run test:streamlit -- --project=chromium

First, you need to run (from the mitosheet/ directory) in this same virtual environment::

jupyter labextension develop . --overwrite

For JupyterLab, run from the tests/ directory:

jupyter lab --config jupyter_server_test_config.py

And then in a separate terminal run from the tests/ directory:

npm run test:jupyterlab -- --project=chromium

For Jupyter notebook, run from the tests/ directory: :

jupyter notebook --config jupyter_notebook_config.py

And then in a separate terminal run from the tests/ directory:

npm run test:notebook -- --project=chromium

First, you need to run (from the mito-ai/ directory) in this same virtual environment:

jupyter labextension develop . --overwrite
jupyter server extension enable --py mito_ai

Then, from the tests/directory, set your OPENAI_API_KEY environment variable:

export OPENAI_API_KEY=<your-openai-api-key>

From the same terminal, run:

jupyter lab --config jupyter_server_test_config.py

And then in a separate terminal run from the tests/ directory:

npm run test:mitoai -- --project=chromium

First, run the following command to install the dependencies from the mito-sql-cell directory:

jupyter labextension develop . --overwrite
jupyter server extension enable mito_sql_cell

Then, run the following command to start the JupyterLab server from the tests/ directory:

jupyter lab --config jupyter_server_test_config.py

And then in a separate terminal run from the tests/ directory:

npm run test:mito-sql-cell -- --project=chromium

Run

python dash-test.py

Then, run the tests with

npm run test:dash -- --project=chromium

Add a --headed flag to see the test run.

See the Galata README.md here: https://github.com/jupyterlab/jupyterlab/tree/master/galata, where it fully documents how to create and record (!) tests.

Writing tests for Streamlit is actually a breeze, using VSCode's Playwright extension. To create new tests:

1. Install the Playwright extenstion for VSCode
2. Use the Record New test functionality from the Playwright extension to create a basic test
3. Copy this into the Streamlit test folder. Then, change the setup for the test to use the utilities from the other tests.

Keep in mind that:

1. Setup is done in a standard function.
2. If you click a button that calls the backend (e.g. makes an edit) you need to use clickButtonAndAwaitResponse.

Note that you can also edit a test with the Record at Cursor functionality.